PH31373: CICS TOR REGION HANG. UNABLE TO LOGON TO REGION SETLOGON HOLD IS TURNED ON

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• You notice your Terminal Owning Region appears hung.
  No users are able to logon to the region, so you took
  a dump and cancelled the region.
  .
  The dump shows the CSA time was current, the region
  is not at maxtask or short on storage, but looking
  at terminal control, we find SETLOGON HOLD is set, which
  explains why no users were able to logon.
  .
  There are also 99 postponed work elements- PWEs
  Also, there were over 2,000 terminals queue on the
  System Error Queue, waiting to be processed.
  The main CSTP looks healthy, but the CSNE task-
  which is the VTAM node abnormal condition program,
  is not healthy.
  CSNE was suspended on ENQUEUE  KCSTRING, since the prior
  day.  The owner of the ENQUEUE was a BMS paging task- CSPG
  This CSPG task was suspended on TSQUEUE xxxxx.
  The owner of this queue was the CSNE task, so we are
  in a deadly embrace.
  Our investigation has shown that this can arise when recoverable
  queues are being used to route a BMS message to multiple
  terminals and more than one terminal has errors that are being
  handled at the same time.
  Additional symptoms: KIXREVDAM
  This is created for case TS004281387.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All CICS Users.                              *
  ****************************************************************
  * PROBLEM DESCRIPTION: Users unable to log onto CICS region.   *
  ****************************************************************
  External symptom is that users are unable to log onto a CICS
  region, but when looking at the state of the CSNE transaction
  it is in a KCSTRING ENQUEUE wait.
  The owner of the KCSTRING ENQUEUE is the CSPG transaction.
  The CSPG transaction is in a SUSPEND for a TSQUEUE.
  The owner of the TSQUEUE is the CSNE transaction.
  This deadly embrace is caused because DFHZNAC (CSNE) issues a
  CANCEL_AIDS_FOR_TERMINAL call. This drives a BMS PURGE, this
  will issue a PUT_REPLACE call to a recoverable TSQUEUE. It gets
  suspended and whilst suspended holding the TSQUEUE, CSPG does
  the BMS purge, and gets the KCSTRING ENQUEUE and then gets
  suspended on the TSQUEUE. Then another CSNE starts to tidy up
  another terminal and this gets suspended trying to obtain the
  KCSTRING ENQUEUE.
  

Problem conclusion

• DFHZNAC has been changed to issue a SYNCPOINT before the call
  to CANCEL_AIDS_FOR_TERMINAL which does the BMS PURGE.
  

Temporary fix

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

• APAR is sysrouted TO one or more of the following:

  UI74152 UI74153 UI74154 UI74155

Modules/Macros

• DFHZNAC
  

Fix information

• Fixed component name

  CICS TS Z/OS V5

• Fixed component ID

  5655Y0400

Applicable component levels

• R000 PSY UI74155

     UP21/02/27 P F102

• R100 PSY UI74154

     UP21/02/27 P F102

• R200 PSY UI74153

     UP21/02/26 P F102

• R300 PSY UI74152

     UP21/02/26 P F102

Fix is available

• Select the PTF appropriate for your component level. You will be required to sign in. Distribution on physical media is not available in all countries.